yeah but that's a way they want you to behave in order to set up a control group within the target group that continues to behave as expected. the questions to be answered are not which parts of that control group, and how, nudge which parts of the target group slightly off the predicted and/or confirmed results. they answered that way back when. the question is, how can we react to the unexpected results that we ourselves forced. they can't just go on doing the opposite of what's good for them and bad for the users or vice versa, they have 50 years of data on that, some of which, should be noted, was accidently burned or bombed with a bunch of incriminating evidence shortly before investigators arrived ... which should make even the last sus person understand, it wasn't on purpose